Building a mine-level database of methane emission factors for China’s underground coal

Abstract

Translate Article

Take Notes

Abstract Methane (CH 4 ) from China’s underground coal mines is a significant warming source and a priority for near-term mitigation. Yet mine-level inventories remain uncertain because publicly available measurements are sparse, outdated, and inconsistent. We assemble a 2011–2023 mine-level database of methane emission factors (EFs) by integrating the national coal-mine gas assessment with provincial disclosures and other sources. We harmonize definitions and impute missing data (10%–30% for key predictors; 23% for relative gas emissions) using a random-forest multiple-imputation framework incorporating geological and operational predictors and propagating uncertainty. The database reveals pronounced spatiotemporal heterogeneity, with a 92% reduction in small mines despite broadly stable national capacity. National production-weighted EFs declined from 5.39 kg t −1 in 2011 to 4.24 kg t −1 in 2020 as de-capacity policies retired high-emission mines and operational upgrades reduced intensities, and then rebounded to 4.78 kg t −1 in 2023, mainly due to higher production in provinces with above-average EFs, with a possible contribution from deeper mining. A persistent north–south gradient remains, and provinces dominated by coking coal exhibit systematically higher emissions. Selective non-disclosure by high-gas mines may introduce non-random missingness and bias the upper tail downward. Uncertainty analysis shows larger absolute but smaller relative errors for high-gas mines, with total uncertainty dominated by model generalization. The integrated dataset enables targeted mitigation at mine and provincial levels and provides a transparent basis for prioritizing high-emission and high-uncertainty regions. It highlights the need for standardized reporting, targeted measurements, and continuous updates to strengthen China’s coal-methane governance and NDC-aligned policy design.